Post-processing Technology And Decoy-state Technology In Quantum Key Distribution

In cryptography, the battle between cryptographers and cryptanalysts has contin-ued for thousands of years. Living in a time when a quantum world will come even-tually, we cannot turn the mission of privacy to those complexity-based cryptography algorithms anymore. No doubt, quantum key distribution (QKD) is the optimal candi-date to fight against powerful quantum computers. Combined with one-time-pad (OTP) encryption, QKD based on fundamental laws of quantum mechanics provides uncondi-tional security in privacy.One important principle in quantum cryptography is that never underestimate your adversary (eavesdropper) who owns powerful equipment and superior intelligence. In other words, your adversary can achieve almost everything without violating the laws of quantum mechanics. In theory, the unconditional security of QKD is guaranteed even when faced with this powerful adversary. However, the practical implementation of an ideal QKD protocol is always not so ideal, which opens a door for the adversary to attack your practical QKD systems. There are two solutions in practical QKD systems, one is to analyze the actual QKD system carefully and give corresponding schemes to make up for the imperfection, and the other is to design device-independent (DI) QKD protocols which do not rely on the detailed implementation. Due to the formidable technological challenges in DI-QKD, measurement-device-independent (MDI) QKD which can im-mune all possible attacks on measurement devices is the focus of researchers. While there still exist other problems in MDI-QKD, so the updated versions of MDI-QKD are becoming a new focus for researchers.This dissertation focuses on my major research results in practical QKD systems, which are arranged as follows:1. A scheme of length-adaptive privacy amplification in QKD, which can be used in high-speed QKD systems, is proposed. We constructed an optimal length-adaptive multiplication algorithm in privacy amplification, and realized the fast implementation of length-adaptive privacy amplification in QKD, which is good at reducing the finite size effect at large input blocks in high-speed QKD systems.2. A scheme of delayed error verification in QKD is proposed. We can get extra keys in privacy amplification implemented by a larger compression factor Toeplitz ma-trix which is usually adopted, to perform error verification. Essentially, error verification is the same as authentication in QKD. Combining error verification and authentication together, we realized a scheme of delayed error verification in QKD. The delayed error verification scheme simplifies the post-processing pro-cedure in QKD, and also reduces the leakage of information during the regular error verification procedure.3. We investigated the decoy-state MDI-QKD protocol based on the Clauser-Horne-Shimony-Holt (CHSH) inequality (hereafter this protocol is referred to CHSH-MDI-QKD). This protocol only requires the quantum state to be prepared in the two-dimensional Hilbert space and Alice’s (Bob’s) encoding device is indepen-dent of Eve’s. Simulation results show that this scheme is very practical with current technology.4. We investigated the decoy-state MDI-QKD protocol with qubit sources based on mismatched-basis statistics (hereafter this protocol is referred to Mismatched-Qubit-MDI-QKD). By exploiting the mismatched-basis statistics in the security analysis, MDI-QKD even with uncharacterized qubits can generate secret keys. Simulation result shows that this scheme is very promising with current technol-ogy.